The Frequency and Degree of “intrusive /r/” in New Zealand English 65- Conclusion
In our logistic regression analysis of the likelihood of /r/-insertion, we found
a strong social class effect, with speakers from higher social classes signi¿ -
cantly less likely to use intrusive /r/. We also found linguistic factors—the
identity of the preceding vowel and the identity of a following af¿ x consider-
ably affect the likelihood of /r/-insertion.
We also conducted a linear regression analysis in an attempt to predict the
F3 of the /r/s that were produced (where F3 is a measure of the degree of con-
sonantal constriction). We included in the model a “baseline” measurement for
each speaker of average F3 measurements for “real” /r/s (from utterances of the
word Sarah). Any further factors retained in the model, then, index the degree
to which an individual’s intrusive /r/ differs from their “real” /r/. We found that
social class is retained in the model, such that speakers from higher social classes
have higher F3s (and so less constriction relative to their “real” /r/) than lower
social classes do. Thus, intrusive /r/ is a gradient sociolinguistic variable, which
differs not only in frequency of occurrence, but also in degree of realization.
In terms of /r/ after /au/, we demonstrated that the likelihood that an indi-
vidual will insert /r/ in this environment is signi¿ cantly correlated with the
degree to which their /au/ vowel is monophthongal (as measured by the aver-
age Euclidean distance (in F1/F2 space) between two target points during /au/
vowels produced in a non /r/-inserting environment). This is signi¿ cant as it
demonstrates that a phonetic change can make an environment eligible for an
ostensibly phonological rule.
These results cast considerable doubt on current phonological analyses which
regard intrusive /r/ as a straightforward, categorical, non-variable phonological
process. Both the frequency and degree of the /r/ are considerably inÀ uenced
by both linguistic and social factors. These results also raise the more general
question of how often “frequency” may also equate with “degree” in other pho-
nological sound changes that have historically been analyzed as categorical.
Appendix 2.1
When Sarah got her ¿ rst bra, she couldn’t talk about anything else. She could only
talk bra-ese.
Bendon plans to completely bra-ify New Zealand
When Sarah got her ¿ rst bra, she couldn’t talk about anything else. Her speech was
just full of bra-isms.